Professional Documents
Culture Documents
Identification of Patients at Risk For Development of Tertiary Peritonitis On A Surgical Intensive Care Unit
Identification of Patients at Risk For Development of Tertiary Peritonitis On A Surgical Intensive Care Unit
DOI 10.1007/s11605-009-0882-y
ORIGINAL ARTICLE
Received: 31 December 2008 / Accepted: 24 March 2009 / Published online: 8 April 2009
# 2009 The Society for Surgery of the Alimentary Tract
Abstract
Background Tertiary peritonitis (TP) is defined as a severe recurrent or persistent intra-abdominal infection after adequate
surgical source control of secondary peritonitis (SP). The aim of this study was to analyze the characteristics of patients with
SP who will further develop TP in order to define early diagnostic markers for TP.
Study Design Over a 1-year period, all patients on the surgical intensive care unit (ICU) with SP were prospectively
assessed for the development of TP applying the definition of the ICU consensus conference. The Mannheim Peritonitis
Index (MPI), C-reactive protein (CRP) and Simplified Acute Physiology Score II (SAPS II) were assessed at the initial
operation (IO) that was diagnostic for SP and in the postoperative period.
Results Among 69 patients with SP, 15 patients further developed TP, whereas 54 patients did not develop TP. Compared to
SP, patients with transition to TP had significantly higher MPI at IO (28.6 vs. 19.8; p<0.001), relaparotomy rate (2.00 vs.
0.11; p<0.001), mortality (60% vs. 9%; p<0.001), duration of ICU stay (14 vs. 4 days; p<0.005), as well as SAPS II (45.1
vs. 28.4; p<0.005) and CRP (265 mg/dL vs. 217 mg/dL; p<0.05) on the second postoperative day after IO.
Conclusions The MPI at IO as well as CRP and SAPS II at the second postoperative day helps to identify patients at risk for
tertiary peritonitis.
Mannheim Peritonitis Index and SAPS II to early identify are expressed as means ± SEM and compared by T test. p
patients with SP at risk for the development of TP. values ≤0.05 were considered statistically significant. To
determine the diagnostic accuracy of the Mannheim
Peritonitis Index measurement during initial operation as
Material and Methods well as C-reactive protein and SAPS II measurements
2 days after initial operation, for the distinction between
Study Population and Definition of Secondary/Tertiary TP and SP, corresponding receiver operating characteristic
Peritonitis curves were calculated. Furthermore, the area under the
receiver operator characteristic curve was defined. Cut-off
During a 1-year period (01.01.2006–31.12.2006), all patients values for the Mannheim Peritonitis Index, C-reactive
admitted to the surgical intensive care unit with a SP were protein, and SAPS II with the corresponding sensitivity,
recorded in a prospective database. Due to hospital policy, all specificity, and confidence intervals were given. Data
patients with a secondary peritonitis are mandatorily admitted were processed with SPSS 16.0/GraphPadPrism 5.
to the surgical intensive care unit—for at least 24 h. SP had to
be diagnosed during a laparotomy, which was referred to as
the IO (Fig. 1). During follow-up, patients with SP were Results
continuously analyzed for the diagnosis of TP—in accor-
dance with the “International Sepsis Forum Consensus Demographic Data of the Study Population
Conference”.2 TP was therefore defined as intra-abdominal
infection that persists or recurs ≥48 h following successful Over a 1-year period (2006), 1,091 patients were admitted to
and adequate surgical source control during the IO.2 As the surgical intensive care unit. Among the 1,091 intensive
indicated in Fig. 1, patients with an obvious failure of care unit patients, 69 were diagnosed having SP. The
surgical source control after the IO or following procedures diagnosis of SP was made intra-operatively in all 69 patients
(e.g., insufficiency of the rectal stump, anastomotic insuffi- during the IO. Among those, 15 patients (21.7%) further
ciency, etc.) were considered as ongoing SP and not as TP. developed TP—according to the ICU consensus conference
Demographic data, origin of peritonitis and intra- definition.2 These patients were referred to as TP patients
operative findings during IO, type of surgical procedure throughout this study. The remaining 54 patients with SP
performed during IO, antibiotic treatment, and follow-up (78.3%) did not develop TP and were therefore referred to as
procedures like relaparotomies were collected. In order to SP patients. Demographic data of the study population are
assess the severity of peritonitis as early as at the IO, the summarized in Table 1. There was no significant difference
Mannheim Peritonitis Index was calculated routinely during in gender distribution, age, body mass index, cardiovascular
the IO as previously described.8,9 Furthermore, C-reactive and pulmonary co-morbidities as well as malignant diseases
protein was monitored daily during the first three postop- between SP and TP patients (Table 1).
erative days and on postoperative day 7 after IO. SAPS II
scores were recorded during the first three postoperative Etiology and Source of Peritonitis
days after IO as previously described.14 Mortality was
defined as any death during postoperative hospitalization. Etiologies and infection source of secondary peritonitis for
Furthermore, intra-operative specimens of abdominal fluid all patients (n=69) found at the IO are depicted in Fig. 2,
were analyzed by standard microbiological techniques. separately for TP patients (n=15) and SP patients (n=54).
The majority of patients had perforated diverticulitis or
Statistical Analysis other colonic perforations at the IO. Less frequent were
other causes like gastric/duodenal perforations, anastomotic
Results for the Mannheim Peritonitis Index were expressed as insufficiencies, or appendicitis. However, there was no
median and displayed in box plots. Box plots are representing significant difference in terms of anatomical site and source
the lower, median, and upper quartile whereas whiskers of infection between TP and SP patients.
indicate the 10th–90th percentile. Outliers are illustrated by
dots. Age, body mass index, Mannheim Peritonitis Index, Detailed Characteristics of TP patients
intensive care unit stay, and the number of relaparotomies
per patient were compared by Mann–Whitney test. Detailed patient characteristics of TP patients are summarized
Frequencies for co-morbidities, underlying malignancy in Table 2. In only one out of the 15 patients (6.7%), TP was
as well as mortality data, frequency of relaparotomies and diagnosed non-operatively. In this patient (patient #14), the
frequency of specific bacteria were compared by Fisher’s diagnosis of TP was made 5 days after the IO by clinical
exact test. C-reactive protein values and SAPS II scores signs of infection and laboratory and CT radiographic
J Gastrointest Surg (2009) 13:1358–1367 1361
Table 1 Demographic Data of Patients with Secondary Peritonitis Who Further Developed Tertiary Peritonitis (TP Patients) and Who Did not
Develop Tertiary Peritonitis (SP Patients)
SP patients TP patients
measurements. In the remaining 14/15 patients (93.3%), TP compared to SP patients (19.8±SD 8.2; median 20, range 4–
was diagnosed intra-operatively by relaparotomies after the 37; p≤0.001, Mann–Whitney test) as illustrated in Fig. 3.
IO (either first or second relaparotomy; Table 2). As required Elevated severity of peritonitis at the IO of TP patients was
by the ICU consensus conference definition of TP, these paralleled by a higher frequency of relaparotomies following
patients showed persistent or recurrent peritonitis ≥48 h the IO (14/15 patients; 93.3%) compared to SP patients (5/54
following successful and adequate surgical source control patients; 9.3%; p≤0.001; Fisher’s exact test; Table 3). The
which was achieved during IO.2 There was no failure of mean number of relaparotomies following IO per patient was
surgical source control of the IO (e.g., insufficiency of the 2.00 (±0.93 SD) for TP patients compared to 0.11 (±0.37 SD)
rectal stump, anastomotic insufficiency, etc.). The median for SP patients (p≤0.001; Mann–Whitney test; Table 3). All
time period between initial operation and diagnosis of TP relaparotomies in the five SP patients were “programmed
was 87 h (range 48–338 h). relaparotomies”. In the TP group, there were nine patients
with “programmed relaparotomies” and five patients with “on
Severity, Clinical Course, and Outcome of Secondary demand relaparotomies” that were initiated by clinical
and Tertiary Peritonitis detection. As a consequence, the concept of “programmed
relaparotomies” was applied with a significantly higher
The mean Mannheim Peritonitis Index, which was recorded at frequency in TP patients (60.0%) compared to SP (9.3%;
the IO in all patients (n=69), revealed significant higher values p≤0.001; Fisher’s exact test; Table 3). The timing and
for TP patients (28.6±SD 7.0; median 20, range 17–39) chronology of relaparotomies in relation to the IO is illustrated
Table 2 Detailed Clinical Data for Patients with Tertiary Peritonitis (n=15)
Patient Age Sex Diagnosis at initial Initial operation (IO) MPI Num. 1st 2nd Diag. ICU f/u (days)
no. (years) operation (IO) at IO relap. relap relap TP (h) stay
(h) (h) (days)
f female, m male, IO initial operation, MPI Mannheim Peritonitis Index, Num. Relap. number of relaparotomies, 1st/2nd relap Time period
between initial operation and first/second relaparotomy in hours, Diag. TP time period between initial operation and diagnose of tertiary
peritonitis (TP), f/u follow-up, Surv. patient still alive, Died (x) patient died x days after the initial operation in the hospital
a
Diagnosis of TP was made based in clinical and laboratory findings only (patient # 14)
day after initial operation was 0.696 (95% confidence leading to high mortality. There is still an ongoing debate
interval=0.562–0.830; p=0.02) and 0.797 (95% confidence about the definition of TP. In fact, some opinions deny the
interval=0.634–0.960; p≤0.001), respectively. A cut-off existence of TP as a distinct entity. In the past, TP has
value for C-reactive protein of 215 mg/L led to a sensitivity simply been defined as failed surgical source control or
of 80.0% and a specificity of 57.4%. A cut-off value of 39 inadequate antibiotic therapy of SP. Other definitions
for the SAPS II score revealed a sensitivity of 80.0% with a emphasized the impaired host response to peritoneal
specificity of 74.5% (Table 4). infection.16 This heterogeneity of definitions resulted in
varying inclusion criteria and incommensurable results in
clinical studies focusing on TP.5 In the current study, we
Discussion applied the latest ICU consensus conference guideline that
provides a precise definition. TP was defined as intra-
Definition of Tertiary Peritonitis abdominal infection that persists or recurs ≥48 h following
successful and adequate surgical source control.2 This
The standard treatment for SP is an immediate laparotomy definition contains two essential conditions, which have to
with surgical source control and antibiotic therapy. How- be met: the time period (≥48 h) and successful surgical
ever, a few patients will develop a clinical syndrome—also source control. Although the ICU guideline does not
referred to as TP, which is characterized by a persistent provide further explanation for “successful surgical source
intra-abdominal infection, an altered microbial flora, failure control”,2 our interpretation of this term was a complete and
of the immune response, and progressive organ dysfunction sustainable eradication of the surgical focus. If a patient
Figure 5 Microbiological
spectrum of microbial isolates
obtained from the initial
operation. Comparison between
patients who further developed
tertiary peritonitis (TP; n=15
specimens) compared to patients
who did not (SP; n=54
specimens). Significantly higher
proportion of E. coli in TP
compared to SP (*p≤0.05;
Fisher’s exact test). Dotted lines
separate gram-negative bacteria,
gram-positive bacteria, and
fungi.
J Gastrointest Surg (2009) 13:1358–1367 1365
presented—during relaparotomy or clinically—with an Several epidemiologic and clinical risk factors have already
obvious failure of previous surgical source control in terms been identified that might predispose to TP, which include
of a “technical problem”, this patient was not classified as age, etiology of peritonitis, malnutrition, and multi-resistant
TP but as SP patient. Other examples of “failure of surgical microorganisms.15 With regard to the patient’s age or
source” control comprise insufficiency of the rectal stump etiology and infection source of peritonitis, we were unable
after Hartmann’s procedure, anastomotic insufficiency, or to detect significant differences between TP and SP.
other technical problems that lead to disruption of the Concerning the microbial flora encountered in the initial
physical integrity of the gastrointestinal hollow organs. operation, we did only find a higher proportion of E. coli in
Nevertheless, there is consensus that SP and TP exist in TP patients compared to SP patients. All other bacteria
a continuum and the transition between both may be quite were equally distributed. It has recently been shown that
subtle. Although TP may be diagnosed during relaparotomy there is a microbial shift in TP towards Enterococcus,
as a simple discrete point in the illness, in reality, it evolves Enterobacter, Pseudomonas, Candida albicans and other
gradually over several hours or days. In the current study, opportunistic bacteria and fungi.11,12,17 However, in this
TP was diagnosed during relaparotomy in 14/15 patients. study, we could only demonstrate a significant shift towards
Only one patient was diagnosed having TP by clinical and Enterococcus and C. albicans between patients who
laboratory measures 120 h after initial operation. For all suffered from TP compared to SP. In our opinion,
protein constitutes a routine parameter in patients with critically ill patient. Curr Opin Crit Care 2001;7:117–121.
doi:10.1097/00075198-200104000-00010.
abdominal infections, it has hardly been explicitly evaluated
6. Koperna T, Schulz F. Relaparotomy in peritonitis: prognosis and
in the diagnosis of TP.5,15 In our study, the time course of C- treatment of patients with persisting intraabdominal infection.
reactive protein displayed a curve with two peaks: one peak World J Surg 2000;24:32–37. doi:10.1007/s002689910007.
preoperatively and one peak on the second postoperative day 7. Koperna T. Surgical management of severe secondary peritonitis. Br
J Surg 2000;87:378. doi:10.1046/j.1365-2168.2000.01369-12.x.
after the IO. In between, on the first postoperative day, lower
8. Linder MM, Wacha H, Feldmann U, Wesch G, Streifensand RA,
C-reactive protein values were observed, possibly due to an Gundlach E. The Mannheim Peritonitis Index. An instrument for
operative clearing effect. Interestingly, although both curves the intraoperative prognosis of peritonitis. Chirurg 1987;58:84–
run parallel, C-reactive protein values of TP patients were 92.
9. Billing A, Frohlich D, Schildberg FW. Prediction of outcome
significantly higher compared to SP patients on the peak of using the Mannheim Peritonitis Index in 2003 patients. Peritonitis
the second postoperative day (265 vs. 217) after the IO. Study Group. Br J Surg 1994;81:209–213. doi:10.1002/
However, the corresponding area under the receiver operator bjs.1800810217.
characteristic curve was only 0.696. The main problem of C- 10. Fugger R, Rogy M, Herbst F, Schemper M, Schulz F. Validation
study of the Mannheim Peritonitis Index. Chirurg 1988;59:598–
reactive protein is the lack of specificity for abdominal
601.
infections, as shown in numerous studies.18–20 A rise of C- 11. Panhofer P, Izay B, Riedl M, Ferenc V, Ploder M, Jakesz R,
reactive protein during the postoperative period may simply Gotzinger P. Age, microbiology and prognostic scores help to
be the result of the operative trauma.21,22 Nevertheless, this differentiate between secondary and tertiary peritonitis. Langen-
becks Arch Surg 2009;394(2):265–271.
study shows a high diagnostic potential of C-reactive protein.
12. Weiss G, Meyer F, Lippert H. Infectiological diagnostic problems
This hypothesis has to be addressed in further studies. in tertiary peritonitis. Langenbecks Arch Surg 2006;391:473–482.
doi:10.1007/s00423-006-0071-3.
13. Agha A, Bein T, Frohlich D, Hofler S, Krenz D, Jauch KW.
“Simplified Acute Physiology Score” (SAPS II) in the assessment
Conclusion of severity of illness in surgical intensive care patients. Chirurg
2002;73:439–442. doi:10.1007/s00104-001-0374-4.
In conclusion, due to high mortality of tertiary peritonitis and 14. Le Gall Jr., Lemeshow S, Saulnier F. A new Simplified Acute
often delayed diagnosis, it is crucial to identify patients at risk Physiology Score (SAPS II) based on a European/North American
multicenter study. JAMA 1993;270:2957–2963. doi:10.1001/
for developing tertiary peritonitis as early as possible: at the
jama.270.24.2957.
initial operation that reveals the diagnosis of peritonitis and 15. Malangoni MA. Evaluation and management of tertiary peritoni-
during the first postoperative days. Our results indicate that tis. Am Surg 2000;66:157–161.
the Mannheim Peritonitis Index assessed at the initial 16. Evans HL, Raymond DP, Pelletier SJ, Crabtree TD, Pruett TL,
Sawyer RG. Tertiary peritonitis (recurrent diffuse or localized
operation and the time course of C-reactive protein and SAPS
disease) is not an independent predictor of mortality in surgical
II during the first days after initial operation are promising patients with intraabdominal infection. Surg Infect (Larchmt)
diagnostic candidates for the future. 2 0 0 1 ; 2 : 2 5 5 – 2 6 3 . d is c u s s i o n 2 6 4 – 2 5 5 . d o i : 1 0 . 1 0 8 9 /
10962960152813296.
Acknowledgments The authors thank all the members of the 17. Nathens AB, Rotstein OD, Marshall JC. Tertiary peritonitis:
intensive care unit team for continuous support. They further thank clinical features of a complex nosocomial infection. World J Surg
Tim Holland-Letz, Department of Medical Informatics, Biometry and 1998;22:158–163. doi:10.1007/s002689900364.
Epidemiology, Ruhr-University Bochum for statistical assistance. 18. Dahaba AA, Hagara B, Fall A, Rehak PH, List WF, Metzler H.
Procalcitonin for early prediction of survival outcome in postop-
erative critically ill patients with severe sepsis. Br J Anaesth
2006;97:503–508. doi:10.1093/bja/ael181.
References 19. Mokart D, Merlin M, Sannini A, Brun JP, Delpero JR,
Houvenaeghel G, Moutardier V, Blache JL. Procalcitonin,
1. Weiss G, Steffanie W, Lippert H. Peritonitis: main reason of interleukin 6 and systemic inflammatory response syndrome
severe sepsis in surgical intensive care. Zentralbl Chir (SIRS): early markers of postoperative sepsis after major surgery.
2007;132:130–137. doi:10.1055/s-2006-960478. Br J Anaesth 2005;94:767–773. doi:10.1093/bja/aei143.
2. Calandra T, Cohen J. The international sepsis forum consensus 20. Reith HB, Mittelkotter U, Debus ES, Kussner C, Thiede A.
conference on definitions of infection in the intensive Procalcitonin in early detection of postoperative complications.
care unit. Crit Care Med 2005;33:1538–1548. doi:10.1097/ Dig Surg 1998;15:260–265. doi:10.1159/000018625.
01.CCM.0000168253.91200.83. 21. Lindberg M, Hole A, Johnsen H, Asberg A, Rydning A, Myrvold
3. Reemst P, Van Goor H, Goris RJ. SIRS, MODS and tertiary HE, Bjerve KS. Reference intervals for procalcitonin and C-
peritonitis. Eur J Surg 1996;576(Suppl):47–48. discussion 49. reactive protein after major abdominal surgery. Scand J Clin Lab
4. Buijk SE, Bruining HA. Future directions in the management of Invest 2002;62:189–194. doi:10.1080/003655102317475443.
tertiary peritonitis. Intensive Care Med 2002;28:1024–1029. 22. Meisner M, Tschaikowsky K, Hutzler A, Schick C, Schuttler J.
doi:10.1007/s00134-002-1383-6. Postoperative plasma concentrations of procalcitonin after differ-
5. Evans HL, Raymond DP, Pelletier SJ, Crabtree TD, Pruett TL, ent types of surgery. Intensive Care Med 1998;24:680–684.
Sawyer RG. Diagnosis of intra-abdominal infection in the doi:10.1007/s001340050644.